Journal of Paleolimnology (2006) 35:441–465  Springer 2006

DOI 10.1007/s10933-005-1995-2

Human impact since medieval times and recent ecological restoration

in a Mediterranean lake: the Laguna Zoñar, southern Spain

Blas L. Valero-Garcés1,*, Penélope González-Sampériz1, Ana Navas2, Javier Machı́n2,

Pilar Mata3, Antonio Delgado-Huertas4, Roberto Bao5, Ana Moreno1, José S. Carrión6,
Antje Schwalb7 and Antonio González-Barrios8
1
Pyrenean Institute of Ecology – CSIC, Apdo 202, E-50080 Zaragoza, Spain; 2EEAD-CSIC, Apdo 202,
E-50080 Zaragoza, Spain; 3Facultad Ciencias del Mar y Ambientales, Geologı´a, Campus Rı´o San Pedro, s/n
11510 Puerto Real, Cadiz, Spain; 4Estación Experimental del Zaidı´n (CSIC), Prof. Albareda 1, 18008
Granada, Spain; 5School of Sciences, University of A Coruña, Campus da Zapateira s/n, E-15071 A Coruña,
Spain; 6Department of Vegetal Biology (Botanic Area), Faculty of Biology, Campus de Espinardo, University
of Murcia, 30100 Murcia, Spain; 7Institut für Geowissenschaften, Technische Universität Braunschweig,
D-38106 Braunschweig, Germany; 8Cátedra de Medio Ambiente ENRESA, Campus Rabanales, Edificio
Aulario Averroes, Córdoba University, 14071 Córdoba, Spain; *Author for correspondence (e-mail: blas@
ipe.csic.es)

Received 11 February 2005; accepted in revised form 1 August 2005

Key words: Ecological restoration, Human impact, Late Holocene, Olive cultivation, Paleohydrology,
Western Mediterranean

Abstract

The multidisciplinary study of sediment cores from Laguna Zoñar (37
29¢00¢¢ N, 4
41¢22¢¢ W, 300 m a.s.l.,
Andalucı́a, Spain) provides a detailed record of environmental, climatic and anthropogenic changes in a
Mediterranean watershed since Medieval times, and an opportunity to evaluate the lake restoration policies
during the last decades. The paleohydrological reconstructions show fluctuating lake levels since the end of
the Medieval Warm Period (ca. AD 1300) till the late 19th century and a more acute dry period during the
late 19th century – early 20th century, after the end of the Little Ice Age. Human activities have played a
significant role in Laguna Zoñar hydrological changes since the late 19th century, when the outlet was
drained, and particularly in the mid-20th century (till 1982) when the spring waters feeding the lake were
diverted for human use. Two main periods of increased human activities in the watershed are recorded in
the sediments. The first started with the Christian conquest and colonization of the Guadalquivir River
Valley (13th century) particularly after the fall of the Granada Kingdom (15th century). The second one
corresponds to the late 19th century when more land was dedicated to olive cultivation. Intensification of
soil erosion occurred in the mid-20th century, after the introduction of farm machinery. The lake was
declared a protected area in the early 1980s, when some agricultural practices were restricted, and con-
servation measures implemented. As a consequence, the lake level increased, and some littoral zones were
submerged. Pollen indicators reflect this limnological change during the last few decades. Geochemical
indicators show a relative decrease in soil erosion, but not changes in the amount of chemical fertilizers
reaching the lake. This study provides an opportunity to evaluate the relative significance of human vs.
climatic factors in lake hydrology and watershed changes during historical times. Paleolimnological
reconstructions should be taken into account by natural resources agencies to better define lake manage-
ment policies, and to assess the results of restoration policies.
442
Introduction
Lakes in the inland areas of Spain have been part
of the cultural landscape for several millennia.
Human activities around the lakes included not
only the use of water, but also other associated
resources as salt minerals, fishing and hunting, and
farming. Some archaeological sites suggest a
significant use of the lake resources and the
watersheds since the Neolithic, with an increase
during Roman times, and particularly in the last
few centuries (Ebro Basin, Davis 1994; Salines
Lake, Giralt et al. 1999; Estaña Lake; Riera et al.
2004). In Mediterranean areas, where water re-
sources are scarce, historical human activities have
been a decisive factor in the lake hydrology and,
particularly, agricultural practices a main forcing
in the depositional dynamics of the lake systems
(e.g., Salada de Chiprana, Ebro Basin; Valero-
Garcés et al. 2000a). On the other hand, most
paleorecords indicate that vegetational and
hydrological changes in the Mediterranean regions
of Spain during the last glacial cycle were re-
sponses to effective moisture crises rather than to
temperature fluctuations (Pons and Reille 1988;
Davis 1994; Peñalba et al. 1997; Giralt et al. 1999;
Valero-Garcés et al. 1998, 2000b, 2004; Carrión
2002; Carrión et al. 2003; González-Sampériz
2004). Paleohydrological tools applied to lake re-
cords may provide the needed rainfall and effective
moisture reconstructions to better understand
climate fluctuations in Mediterranean Spain, and
also contribute to unravel the relationship between
the different cultures with the environment in the
past. During the last 20 years, government agen-
cies have tried to implement plans for lake and
watershed management in many of the Spanish
wetlands. In most cases, little was known of the
lake dynamics and the previous history of the
system, and consequently neither the ‘‘pristine’’
stage to be preserved nor the results of the con-
servation efforts could be evaluated. In these cases,
the study of the sedimentary record bring the only
consistent way to establish targets to be achieved
in any management plan to restore those natural
conditions (Hutchinson 2005; Cohen et al. 2005).
Climate and human impact are main factors
controlling lake dynamics, but they have played a
variable role through times, particularly in his-
torical periods, and it is usually difficult ascribing
specific limnological changes to specific human
activities (Valero-Garcés et al. 2000a; Bradbury
et al. 2004; Kienel et al. 2005). Paleolimnological
studies have shown the effects of intensified ero-
sion caused by land clearance in small (Hutchinson
2005) and large (Cohen et al. 2005) lakes, the
development of hypereutrophic conditions
(Bradbury et al. 2004) and the effects of tour-
ism development and industrialization (Tylmann
2005). Reconstruction of environmental, anthro-
pogenic and climatic forcings during historical
times is also problematic because of the difficulties
of obtaining an accurate chronology.
Here, we present results of paleolimnological
investigations at Laguna Zoñar, the deepest (up to
15 m) lake in the lowlands of Andalucı́a (Spain).
The lake is spring-fed and has no surface outlet,
making it a good candidate for paleohydrological
reconstructions. The lake is located in the Campiña
Cordobesa, a region south of Córdoba with a long
history of agricultural practices and currently
dedicated almost exclusively to the cultivation of
olive trees. Archaeological and historical records
suggest that the lake and its watershed have been
affected by human activities since Roman times
and particularly in the late 19th and 20th centuries
(Montilla Archaeological Museum Archives). In
the 1960s, the waters from the springs were diverted
for human use, lake level lowered, and the outlet
creek became non-functional. Since the area was
declared Natural Park in the early 1984, some
conservation measurements have been imple-
mented, lake level has recovered, and some vege-
tated littoral areas have been submerged. The rapid
response of the lake system during the last decades
to hydrological fluctuations and to human impact
in the watershed provides an opportunity to assess
how decadal to centennial hydrological and envi-
ronmental changes are archived in lake sediments.
Paleolimnological studies in Laguna Zoñar allow
an assessment of the interaction between humans
and the environment against a background of cli-
mate variability, and help to define conservation
policies in Mediterranean wetlands and lakes.
Study site
Physical setting
The Laguna Zoñar is located in the southern
margin of the Guadalquivir River Basin, close to

the Sub-Betic tectonic domain of the Betic Range
(IGME 1988) (Figure 1). Marine formations (yel-
lowish sandstones, siltstones, marls and calcare-
nites) were deposited over the allochthonous
sub-betic formations during the Upper Miocene.
Triassic rocks composed of carbonates, claystones,
evaporites and igneous rocks (ophiolites) outcrop
along the faults. Karstic processes affecting the
limestone and evaporite formations resulted in
large and shallow depressions filled with fine
clastic materials and reddish soils. Permanent and
ephemeral lakes are located in larger depressions,
likely related to faults. Laguna Zoñar, Rincón and
Amarga are permanent brackish and saline lakes;
Laguna Salobral, Tiscar and Jarales are ephemeral
saline lakes.
The southern area of the Cordoba province has
a semi-humid Mediterranean climate with about
Figure 1. Geographic and Geological setting of Laguna Zoñar. (a) Annual Rainfall distribution in Spain (Capel Molina 1981) and
location of the Laguna Zoñar (Z). (b) The Laguna Zoñar watershed. The main three spring are from SW to NE: Eucaliptus, Zoñar and
Escobar. (c) Geological map of the Zoñar Lake (IGME 1988).
443
538 mm of annual rainfall (range 1100–300 mm)
(Figure 1a) and an average temperature of
16.1
C. Over 70% of the precipitation occurs in
late fall and early winter (November–December).
Evapo-transpiration is estimated between 1500
and 1750 mm yr 1 (Enadimsa 1989).
Laguna Zoñar (37
29¢00¢¢ N, 4
41¢22¢¢ W, 300 m
a.s.l.) is the deepest (up to 15.4 m) and largest
(37 ha, surface area) of the 10 lakes that belong to
the Natural Park of the Southern Córdoba (Fig-
ure 1). Its origin has been related to tectonic,
karstic and diapiric activity (Moya 1984; Sánchez
et al. 1992). The elongated lake shape, following
the dominant regional tectonic direction (N 45
–
50
E), indicates a clear tectonic control. On the
other hand, the location close to Triassic outcrops
stresses the importance of diapirism in the genesis
of these structures. Finally, an open doline has
444
been mapped between the mouths of the Lobo and
Moro creeks to the south and east of the lake
(IGME 1988), raising the possibility of a karstic
origin at least for some areas of the Zoñar Basin.
The lakes and wetlands comprise unique envi-
ronments where some endangered species as the
White-headed Duck (Oxyura leucocephala) live.
Aquatic submerged vegetation is dominated by
Naja marina and Zannichellia pallustris. A wide (5–
15 m) area with littoral emergent vegetation
(Phragmites australis, Typha domingensis) sur-
rounded the lake before the early 1980s. The
southeastern depression close to the mouth of the
Lobo Creek is totally occupied by littoral vegeta-
tion.
Limnology and hydrology
The estimated lake water volume ranged between
2.8 · 106 m3 during the 1980s (maximum water
depth was 15.7 m) to 1.9 · 106 m3 in the early
1990s (maximum water depth was 12.3 m) (de-
tailed references in Enadimsa 1989 and Valero-
Garcés et al. 2003). Waters are saline (2.4 g l 1),
alkaline (alkalinity values in surface waters be-
tween 3–5 mg l 1; pH between 7.1 and 8.4) and
dominated by (Cl ) – (SO42 ) and Na+ (Table 1).
The lake is monomictic with a water temperature
range between 9 and 27
C and an average Secci
Disc depth of 3 m. It shows a thermocline and
oxycline at about 4 m depth between May and
September (anoxic below 6 m), and it is mixed in
winter (November–February).
Several ephemeral creeks enter the lake, but the
main water input are the Zoñar, Escobar and
Eucaliptus springs located in the southern margin
of the lake. (Figure 1b). The Zoñar and Escobar
Table 1. Main limnological parameters of Zoñar Lake.
Lake Surface 37 ha
Watershed 876,78 ha
surface area
Max. depth 15.4
Salinity 1–2.4 g l 1
E.C. 2.5–3 mS cm 1
Water type Cl–(HCO3)–(SO4)//Na–(Mg)–(Ca)
Alkalinity 3–5 mg l 1 (surface)
pH 8–9 (surface), 7.4–7.8 (bottom)
Chemocline May–September (3–6 m water depth)
Oxygen Anoxic bottom from May to September
springs are the most important (average flows,
3.5 l s 1 and 1 l s 1, respectively). Springs waters
are (HCO3 ) – (Ca2+), while lake waters are (Cl )
– (Na+). Currently, Laguna Zoñar does not have
a surface outlet, although historically the Arroyo
de las Salinas in the northwestern corner of the
lake was functional and evacuated the water to the
Cabra River and into the Genil River watershed.
This outlet connected with a smaller, shallower
lake (Laguna Chica) and some saline wetlands that
occupied the valley.
The input to the lake during an average year
(Enadimsa 1989) is as follows: rainfall, 0.177 Hm3;
runoff, 0.168 Hm3; groundwater, 0.4 Hm3;
springs, 0.073 Hm3. The only output is through
evaporation, estimated as 0.8 Hm3 per year. The
surface area of the Lobo and Moro creeks water-
sheds are about 9.64 km2 (Enadimsa 1989). Sev-
eral small watersheds (8.23 km2) are located over
aquifer formations draining into the lake, and they
are also part of the hydro-geological watershed of
the lake. The main aquifers are the Miocene cal-
carenites and the Quaternary alluvial deposits, and
the Laguna Zoñar is the discharge area for both
aquifers. A hydrological survey during the year
1984–1985 (Moya 1986) and the data collected by
the regional government since 1982 show that both
lake level and spring flows quickly respond to
rainfall during the hydrological year and also at a
decadal time scale. During the prolonged dry
period of 1992–1995, lake level dropped to a
minimum of 11 m (Figure 2), but quickly recov-
ered after the humid 1996–1997 years. The isoto-
pic composition of the lakewaters plots to the right
of dD- d18O meteoric water line (Valero-Garcés
et al. 2003), suggesting that evaporative processes
play a significant role in the lake hydrology.
Methods
The Laguna Zoñar watershed was identified and
mapped using topographic and geological maps.
Water chemical composition, lake level fluctua-
tions and changes in land uses monitored since
1982 by Andalucian Government Environmental
Agencies were compiled. A seismic survey was
conducted with a 3.5 Khz seismic profiler in June
2002. Unfortunately, sediment penetration was
extremely poor and only bathymetry and bottom
basin morphology could be reconstructed (Valero-

Figure 2. Relationship between annual water input (rainfall and spring flow) and lake level (maximum and minimum) during the
period 1985–2000 (Data from the Environmental Agency of the Andalucian Government).
Garcés et al. 2003). Low seismic penetration is
commonly caused by the presence of gas, and in
Zoñar Lake it could result from decomposition of
the abundant organic matter, both transported
from the watershed and originated in the lake.
Two sediment cores (ZON-01-1A: 1.72 m long,
and ZON-01-1B, 1.17 m long) were retrieved in
the deepest area of Laguna Zoñar (14.5 m water
depth). The 1.17 m long core was sub-sampled in
the field at 0.5 and 1 cm intervals for 210Pb and
137
Cs dating. Magnetic susceptibility was mea-
sured with a Bartington magnetic susceptibility
bridge every 2 cm only in core ZON-01-1A. The
ZON-01-1A core was split in two halves and sed-
imentary facies were defined by macroscopic visual
description including color, grain-size, sedimen-
tary structures, fossil content, and by microscopic
smear slide observations (Schnurrenberger et al.
2003). The core was sub-sampled for organic
matter, carbonate, grain size, mineralogy, trace
element geochemistry, pollen, diatoms, and ostra-
codes. Organic matter content was determined by
loss-on-ignition analyses at 450
C (Heiri et al.
2001) and carbonate content with a Barahona
calcimeter (CSIC 1976). Although LOI is not an
accurate stand-alone technique for the estimation
of the carbon content (Boyle 2004; Santisteban
445
et al. 2004) it provides a useful characterization of
the sediment composition and the sedimentary
facies. Whole sediment mineralogy was charac-
terized by X-ray diffraction, and relatively mineral
abundance was determined using peak intensity.
Grain size was determined using a Coulter particle
size analyzer (Buurman et al. 1997). Samples were
treated with 10% hydrogen peroxide in a water-
bath at 80
C to eliminate the organic matter, a
dispersant agent was added and ultrasound treat-
ment was used prior to measurement.
Bulk sediment samples (0.5 g) were digested
with HF (48%) in microwave (Milestone
1200 mls). Analyses for main elements composi-
tion were performed by atomic emission spec-
trometry using an inductively coupled plasma
ICP-OES with solid state detector (Perkin Elmer
Optima 3200 DV).
Pollen grains and spores were sampled every
10 cm and extracted in the laboratory by the
classic chemical method (Moore et al. 1991) using
Thoulet dense liquid (2.0) for palynomorph con-
centration) and Lycopodium clavatum spore tablets
to calculate pollen concentration. Pollen diagrams
were constructed using Tilia, TiliaGraph and Co-
rel Draw software. Ostracode valves were sepa-
rated from 2-cm-thick slices taken every 20 cm,
446
following the procedure described by Forester
(1988). Samples for diatom analysis were taken
every 10 cm and prepared following standard
procedures (Renberg 1990). At least 400 valves
were counted per sample at 1000· using a Nikon
Eclipse 600 microscope with Nomarski differential
interference contrast optics. Raw diatom counts
were converted to percent abundances.
The chronology is constrained by two AMS 14C
dates from the longer core ZON-01-1A
(593 ± 3814C yr BP at 124–126 cm depth, and
1771 ± 3814C yr BP at 166–167, Table 2) analyzed
at the Arizona Dating Facility and by 210Pb and
137
Cs dating in the parallel shorter core (ZON-01-
1B) performed at the St. Croix River Station
(University of Minnesota). Both cores were cor-
related using sedimentary facies, grain size, and
organic matter profiles (Figure 3).
Results
Chronology
The 210Pb activities in the measured samples
range between 0.384 and 1.658 pCi g 1 and
fluctuate irregularly throughout the core (Fig-
ure 4). There is no discernable down-core trend
and dates cannot be reliably modeled from the
Zoñar 210Pb data. The low 210Pb activities values
are somehow unexpected in a semi-humid region
(average annual precipitation more than 500 mm)
and it could be related to dilution of the atmo-
spheric 210Pb by high amounts of eroded soil.
There is measurable radio-caesium to a depth of
75 cm in core ZON-01-1B with a clear peak at
70–71 cm. Although given the high variability in
sedimentation rates it is not possible to exclude
137
Cs percolation, we feel confident that the peak
represents the 1963 horizon. The cumulative mass
of sediment overlying this 1963 peak indicates a
Table 2. AMS results for core ZON-01-1A.
Sample Material Lab number d13C
124—126 cm Aquatic AA47855 28.0
macrophytes
166–167 cm Pollen AA60921 27.1
concentrate
very high rate of burial, which accounts for the
low and irregular 210Pb activities due to dilution
of the atmospherically-derived 210Pb by eroded
soil material. The sharp drop-off in 137Cs below
70–71 cm in Zoñar is also due to the lowest
concentration of 137Cs below the 1963 peak,
according to the global pattern of the radioiso-
tope fallout during the early period of atmo-
spheric nuclear testing. Another factor in
addition to the lower 137Cs inputs was dilution
from high erosion during this period in which
soils had not yet become enriched in 137Cs. A
likely date of 1963 at 70–71 cm in core ZON-01-
1B represents a mean sedimentation rate of
1.5 cm/yr 1 or 1.0 g cm 2 yr 1 during the last
half century.
The 14C AMS dates (593 ± 3814C yr BP from
aquatic macrophyte remains at the 124–126 cm
depth, and 1771 ± 3814C yr BP from pollen con-
centrate at 166–167 cm interval) indicate that the
Zoñar core ZON-01-1A covers almost the last
2000 yrs and suggest the presence of a hiatus be-
tween the dated intervals (Figure 5). Sedimento-
logical data support the presence of a hiatus at
140 cm depth where an erosive surface is evident.
Based on the age of the upper sample (about AD
1350), and the 1963 horizon correlated between
both cores, the average sedimentation rate for that
interval (Sedimentary Units 3 and 2, see below) is
about 1.6 mm yr 1. The 210Pb and 137Cs analyses
in the parallel core ZON-01-1B indicate a much
higher sedimentation rate for the upper Unit 1
(between 0.7 and 1.8 cm yr 1). This fits with the
assessment of large fluvial input from the Lobo
and Moro Creek during the last decades. The
different thickness of the upper Unit 1 in core
ZON-01-1A (25 cm) and core ZON-01-1B (70 cm)
(Figures 3, 4 and 5) suggests an uneven sedimen-
tation rate over the basin, characteristic of lim-
nic systems dominated by fluvial processes.
Sedimentation rates for laminated facies (Unit 2)
14
C fraction Fm error 14
C yr BP 14
C yr Calibrated
Modern Error year BP(2r)
0.9288 0.0044 593 38 537–651
0.8021 0.0038 1771 38 1591–1817
 447

cm ZON-01-1B ZON-01-1B
0

20

Units ZON-01-1A cm
40 0
A
1 ZON-01-1A
10
B
60 20
C
A
30
B
2 C
D
80 40
E
F 50
G
100 60
A
70

120 80
0 20 40 60 80 100 0 20 40
Grain Size % O.M. 90
3 B
100

110

120

130

140
4 A
150
B
160
C
170
0 20 40 60 80 100 0 20 40
Grain Size % O.M.

Figure 3. Correlation between the two cores ZON-01-1B and ZON-01-1A using sedimentary facies, grain size, and organic matter
profiles.

are likely to be smaller than for massive facies of Sedimentology

Unit 1 where clastic input is higher.
The age model including the 137Cs-derived
chronology for the upper Unit 1 and the 14C AMS
dates is coherent with the pollen data. The sharp
increase in Olea pollen percentages between 30 and
40 cm in core ZON-01-1A (Figure 6) and between
75 and 100 cm in core ZON-01-1B (Figure 4)
marks the Olea rise horizon due to the increase in
olive tree cultivation in the region during the late
19th century.
The 170 cm long core (ZON-01-1A) is composed
of decimeter-bedded, massive, brownish and gray
sediments with an intercalated (24–59 cm depth)
interval composed of finely laminated, variegated
sediments (Figure 5). Sediments are mostly
composed of silicate (quartz and clays) and car-
bonate (calcite) grains, lacustrine and terrestrial
organic matter and biogenic particles (diatoms,
ostracodes). Eight sedimentary facies have been
448

x
ile
us
CORE ZON-01-1B

rc
ue
Units Olea

Q
0 2000
(cm)

20 1990
1a

40 1980
Year
1b
60
1970
1c

1960
80
2

1950
100 0 200 400 600 800 1000
3 Rainfall (mm)
120
0 20 40 60 80 100 0 20 40 20 40 60 0 1 2 3 4 5 0 1 2 20 20 40 60
Grain Size % O.M. % clay minerals Cs-137 Activity Pb- 210 Activity Pollen (%)
(pCi / g) (pCi / g)
<2µ
2 - 50 µ
> 50 µ

Figure 4. Sedimentary Units, sediment composition indicators (grain size distribution, organic matter and clay mineral content),
selected pollen taxa, and profiles of 210Pb and 137Cs age from core ZON-01-1B. Sedimentary Units are correlated with those described
in core ZON-01-1A. The inset shows the annual rainfall in the nearby Aguilar Meteorological Station since 1950s.

identified after integration of visual description,
microscopic observation, grain size, and sediment
composition analyses (LOI and XRD) (Table 3).
Facies 1–5 are massive to faintly laminated car-
bonate mud with variable composition and bed-
ding. Facies 6–8 are organic-rich, finely laminated
facies. Facies are described in detail in Table 3 and
in Valero-Garcés et al. (2003).
The eight facies in the Zoñar core group in two
Facies Associations. Facies Association I inte-
grates cm- to dm- bedded, massive to slightly
laminated calcite muds (F 1, 2, 3 and 4). Facies
Association II integrates finely laminated (F 7 and
8), organic-rich (F 6) and cm-thick, massive, cal-
cite mud (F 5) facies. Facies Association I repre-
sents deposition in Laguna Zoñar during periods
of variable, but significant clastic input. The
absence of lamination in the sediments indicates
intense bioturbation activity, and likely frequent
oxic conditions in the water column. Although
salinity and lake level during deposition of these
calcite mud could vary, the lacustrine system
remained a freshwater lake. The higher siliciclastic
content, the presence of macrophyte rests and
intraclasts, and the erosive nature of the lower
contacts in some layers indicate depositional
conditions for Facies 4 characterized by higher
energy and higher fluvial input. These conditions
would occur in the littoral environments of
Laguna Zoñar and also during flooding episodes
that could reach the deepest areas of the lake.
Gray facies represent mixed lacustrine and alluvial
deposition in the central areas of the lake during
flooding episodes. Brown layers, with higher car-
bonate and diatom content and presence of os-
tracodes, would point to more littoral conditions,
and, consequently, to lower lake levels. Lake level
fluctuated during deposition of facies association I
at relatively higher stand, close to today’s levels.
Facies Association II represents deposition in
Laguna Zoñar during a period of lower clastic
input, and chemical and limnological conditions
more conducive to development of benthic bac-
terial–algal communities. Preservation of fine
laminations indicates absence of bioturbation,
most likely provoked by reduced oxygen in the
bottom waters. The irregular nature of the lami-
nation suggests laminae are not varves. Varie-
gated, organic-rich laminated facies occur in
many brackish-saline lakes where conditions are
more suitable for algal–bacterial communities
than for other lacustrine biota and where anoxia
Figure 5. Sedimentary facies and Units, magnetic susceptibility, sediment composition indicators (grain size distribution, organic matter and carbonate content), selected
geochemical profiles (values in ppm) and 14C AMS dates from core ZON-01-1A.
449
450
Table 3. Sedimentary facies in Zoñar cores.
Facies
Calcitic facies
Facies 1. cm-bedded, massive to
faintly laminated, brownish calcite
mud.
Facies 2. Laminated (1–5 mm),
dark and light brownish calcite
mud.
Facies 3. Massive to faintly lami-
nated gray calcite silty mud.
Facies 4. cm-to dm-bedded, faintly
laminated to massive, dark gray
silty calcite mud with organic
matter.
Facies 5. cm-thick, massive gray
calcite silty mud.
Organic facies
Facies 6. cm-thick, massive, dark
brown organic ooze.
Facies 7. Finely laminated (about
1 mm thick), variegated (gray,
brown, white, green) organic ooze
and calcite mud.
Facies 8. Irregularly laminated
(>2 mm thick) variegated (gray,
brown, green) organic ooze and
calcite mud.
is facilitated by chemical water stratification:
Lake Bogoria, Africa (Renaut and Tiercelin
1994), Mono Lake, USA (Newton 1994), Lake
Hayward, Australia (Coshell and Rosen 1994),
Salada Chiprana, Spain (Valero-Garcés et al.
2000a) are some examples. Water depth for this
Description
These are cm-thick (Units 1 and 4) and dm-thick (Unit 3) layers with gradational
boundaries. O.M. <10%, Quartz <10%. Presence of relatively large (40 · 10 microns)
charcoal particles, and oxidized organic matter particles (soil-origin). Depositional su-
benvironment: sublittoral to distal lacustrine.
Layers are less than 5 cm thick, with sharp, planar contacts. Microfacies similar to
Facies 1. Fine-grained (<10 microns) calcite is commonly dominant. Quartz and sulfide
contents are lower than in Facies 1. Depositional subenvironment: distal lacustrine.
These are cm- to dm-thick layers. Grain size coarser than Facies 1 and 2. Quartz grains
may be larger than 100 microns. Depositional subenvironment: sublittoral lacustrine
with alluvial influence.
Dark gray layers are between 5 and 10 cm thick and they commonly show sharp
boundaries with the underlying beds and more gradational with the overlying beds. Both
calcite grains populations are present, although the coarser grains commonly dominate.
Relatively high quartz content (up to 15%). Both carbonate and quartz grains are
coarser than in other facies. Sulfide content is the highest of the massive facies (1–4).
Organic matter remains showing cell-structures are common and also opaque, oxidized,
organic matter fragments. Depositional subenvironment: distal to sublittoral with allu-
vial influence.
Massive, cm-thick layers (1–2 cm thick) with sharp upper and lower boundaries. They
are composed of homogeneous dark gray silty calcite mud. Quartz content is relatively
higher than in previous facies, although never more than 10%. Sulfide content is high
(up to 15%). Depositional subenvironment: flood deposit in distal areas.
These are cm-thick layers with irregular, erosive lower boundaries and sharp, planar
upper boundaries. Homogeneous, greenish–brownish, amorphous remains of aquatic
origin (algal). Diatom content is also high, up to 15–20%. Calcite content is relatively
low (about 30%) and mostly composed of rice-shaped, euhedral, about 10 microns long
crystals. Depositional subenvironment: benthic algal mat in distal areas.
Four different types of laminae: gray, brown, green and white. Organic matter content is
high (20–30%), and quartz is low (<10%), although always present. Sulfide content is
variable (5–15%), higher in the darker gray laminae. Small (<10 microns), rice-shaped
crystals dominate the calcite components. Green laminae are mostly composed of dia-
toms and the sediment type ranges from diatom ooze to carbonate diatomaceous mud.
Brown laminae show large composition variability from organic (algal and diatoma-
ceous) ooze to organic, diatomaceous calcite mud. Whitish to light gray laminae are fine
calcite mud (euhedral crystals between 5–10 l long). Gray laminae are calcite mud with
higher content of coarser calcite grains, and lower diatom and organic matter content.
Clay minerals dominate over quartz grains. Depositional environment: brackish to saline
lake with benthic algal mat.
Similar to facies 7, but laminae are thicker and more irregular in thickness and the nature
of contacts varies from sharp to diffuse, and sometimes crenulated. Gray and brown
laminae dominate. Depositional subenvironment: brackish to saline lake with benthic
algal mat.
laminated facies ranges from playa-lake settings
to relatively deep meromictic lakes. In Zoñar
sediment core, diatom blooms and ‘carbonate
whitings’ events are registered as thin green and
white laminae, respectively. Coarser grain size in
the laminated intervals from Unit 2 is caused by

the abundance of diatoms (Figure 5). Rare
flooding episodes would deposit thin clay-rich
gray laminae (Facies 5). Clay mineral content is
higher in these laminated, organic-rich facies than
in the massive calcite mud (Figure 4) suggesting
lower energy and deposition out of suspension.
Relatively higher salinities and lower lake levels
could be expected in Laguna Zoñar during
deposition of this facies association.
Four main units have been identified in core
ZON-01-1A (Figure 5): Unit 1 (0–24 cm) com-
posed of cm-bedded brown and gray sediments;
Unit 2 (24–59 cm) composed of laminated sedi-
ments; Unit 3 (59–140 cm) composed of decime-
ter-bedded gray and brown sediments, and Unit 4
(140–170 cm) composed of cm-bedded gray and
brown sediments. The boundaries between the
upper three units are defined by the occurrence of
fine (<1 mm) lamination in Unit 2; the limit be-
tween Units 3 and 4 is marked by the occurrence of
an irregular, erosive surface. A number of subunits
have been defined according to the facies (Figure
5 and see also Valero-Garcés et al. 2003 for a de-
tailed stratigraphy). The upper three units have
been correlated between both cores using sedi-
mentary facies, and sediment composition (organic
matter contents and grain size). Facies association
II only occurs in Unit 2 and facies association I in
Units 1, 3 and 4.
Geochemistry
The chemical composition profiles (Figure 5) show
more constant values in the lower Units 4 and 3
and higher variability in the upper Units 2 and 1.
Aluminum and iron contents are relatively con-
stant in Units 3 and 4 (Fe: 1–1.2%; Al: 1.5–2.5%),
slightly decrease at the base of Unit 2 (Fe: 0.5%;
Al:0.7%) and start an increasing trend at 50 cm
(base of subunit 2 F) till almost the top of Unit 1
(Fe: 1.8%; Al: 2.7%). The top sample shows a
slight decrease in both, aluminum and iron. Alu-
minum reflects the Al-silicates content of the sed-
iments, so it can be considered as a proxy for
alluvial input. Iron is also an indicative of soil
erosion in the watershed, but it also may be ad-
sorbed on clays, or precipitated as colloids and
oxides. Both profiles are coherent with increasing
erosion in the Zoñar watershed during deposition
of Unit 2 and 1, and a small decrease in the last
451
years. Peaks in Fe concentration coincided with
peaks in the Fe/Mn ratio, which has been inter-
preted in other lakes as an evidence of changes in
supply from the catchment more than changes in
the redox conditions in the lake (Boyle 2001).
Potassium and lithium show constant composi-
tions through the sequence except in the organic,
laminated facies of Unit 2. Metals such as zinc and
copper used in fertilizers also show constant values
in Units 4 and 3 and increasing values in Units 2
and particularly 1 (almost double for Zn and
threefold for Cu). No signs of lowering content of
those elements in the sediments occur at the top
sample. Manganese values peak at the laminated,
algal mat intervals and coincide with lower values
of the Fe/Mn ratio. Higher Mn concentrations in
the sediments seem related to strong redox gradi-
ents at the sediment/water interface, as those
expected during deposition of the laminated mats.
Phosphorous concentrations are higher in the
laminated facies of Unit 2 and Unit 1, and their
peaks correlate with intervals of higher organic
matter content. Lower values of Sr in the upper
Unit 1 could be explained by dilution due to
increased input of low Sr marine sediments
(Miocene calcarenites).
Ostracods
The ostracode assemblage from Zoñar (ZON-01-
1A) is characterized by individuals of Ilyocypris
sp., a few Candona sp. and Potamocypris sp. This
assemblage suggests an environment with shallow
and flowing water likely affected by stream input.
The number of ostracode valves in the sediment
samples is very low, and some samples lacked any
ostracods. The sample from the top sediments
(Unit 1: 2–4 cm) contains only Ilyocypris sp.,
suggesting significant stream input into the lake.
Samples from Unit 2 (22–24, 42–44 cm intervals),
the top of Unit 3 (62–64 cm interval) and another
sample from a faintly laminated interval in Unit 3
(122–124 cm) are devoid of ostracodes. The pres-
ence of authigenic calcite in these intervals suggests
that the absence of ostracodes is not due to
carbonate dissolution processes. Most likely,
restricted water circulation and low oxygen con-
tents in the deep areas of the lake as indicated by
laminated sediments impeded survival of benthic
ostracodes. Most individuals in the remaining
samples from Units 3 and 4 are Ilyocypris sp.
452
(102–104, 142–144 cm), indicating oxic bottom
conditions and significant stream input into the
lake prior to deposition of Unit 2. Candona sp. is
only present at 82–84 cm, which could indicate
relatively deeper waters with more restricted cir-
culation. Finally, the sample from Unit 4 (162–
164 cm) is dominated by Potamocypris sp. indi-
viduals, a nektic ostracode that is typical of littoral
environments.
Pollen
Pollen assemblages in samples from both cores are
typical of Mediterranean landscapes dominated by
Olea europaea, evergreenQuercus and Cupressa-
ceae. Olea percentages are the highest in the top
samples (up to 80%) reflecting the large increase in
olive tree cultivation since the late 19th century.
The relative abundances of pollen taxa in core
ZON-01-1A are shown in Figure 6. A few samples
were analyzed in core ZON-01-1B (Figure 4) to
correlate both cores. Three pollen zones have been
identified based on the pollen assemblages.
Pollen Zone ZO-III (160–140 cm). This zone
corresponds to sedimentary Unit 4. Terrestrial
plants are dominated by Mediterranean trees and
shrubs (50–60%): evergreen Quercus, Pistacia,
Rhamnus, Thymelaea, Lycium, Cistus, Ericaceae,
Ephedra, Fabaceae (Genisteae), Lamiaceae,
Hedera helix, etc. Olea is also present, but in lower
percentages than in the upper zones. The sub-
merged Myriophyllum spicatum reaches the maxi-
mum abundances in this zone, suggesting higher
lake levels and fresh waters. The abundance of
Botryococcus and the occurrence of colonies of
Pediastrum also points to relatively high lake lev-
els. Algal spores and spiny acritarch show the
highest percentages and fungal remains the lowest
in the whole sequence. The presence of Zigne-
mataceae zigospores, Gloeotrichia sheats, Rivularia
heterocyst, the fungal spore types Pluricellaespor-
ites and Dyadosporites (Carrión et al. 2001), and
the microfossils 179 and 989 (Carrión and van
Geel 1999) indicate a mesoeutrophic aquatic
environment.
Pollen Zone ZO-II (140–35 cm) includes sedi-
mentary Units 3 and the lower part of Unit 2. The
AP/NAP ratio slightly increases. The low Pinus
percentages suggest long distance input. Terrestrial
vegetation is similar than the previous zone, al-
though a growing trend in Olea suggests increasing
cultivation at a regional scale or just around the
lake. Myriophyllum percentages remain low and
Tamarix disappears in the lower part of this unit.
Hygrophytes abundance and variety increase
(Cyperaceae, Juncus, Typha, Sparganium). In the
upper part of pollen zone II, corresponding to
sedimentary Unit 2, Myriophyllum decreases until
disappearing at the top of the unit, and Sparga-
nium increases suggesting expanding littoral areas.
Algal spores are still well represented, although in
lower percentages than in zone III. This unit is
characterized by the highest percentages of fungal
spores and other palynomorphs: Palaeomycites,
Monoporisporites, Chaetomium, Diporisporites,
Dyadosporites, Pluricellaesporites, Didimoporis-
poronites, Dicellaesporites (Carrión et al. 2001).
These assemblages represent increasing littoral
areas colonized by plants and also an increase in
trophic conditions in the lake, likely with an in-
crease in particulate organic matter in the water.
Pollen Zone ZO-I (35–0 cm) corresponds to the
upper sedimentary Unit 2 and Unit 1. Olea is the
dominant taxa, which reflects the large increase in
olive tree cultivation since the late 19th century
(Ortega Alba 1975) The typical Mediterranean
component (Cupressaceae, evergreen Quercus,
Lycium, Cistus, Ephedra, Genista-Adenocarpus,
etc.) is well represented. Aquatic macrophytes taxa
decrease and eventually Myriophyllum and Pota-
mogeton disappear. However, Potamogeton and
Tamarix appear again at the top of the sequence,
suggesting a relative fresher waters and an increase
in lake level during modern times.
In pollen zones III and II, indicators of human
activity are common: Rumex, Plantago, Vitis, Ci-
chorioideae, Chenopodiaceae, Artemisia, Centau-
rea, Urticaceae, and also Cerealia type. In pollen
zone I, most of these taxa disappear. Algal spores
are also greatly reduced. Fungal spores and
Ascomycete sporocarps are abundant, which sug-
gests that eutrophic conditions are similar to those
of the previous pollen zone II. However, eutrophic
conditions were not extremely high, because the
fungal spores percentages are relatively small
compared to other sites (Carrión and Navarro
2002). The presence of Pseudoschizaea cysts has
been interpreted as an indicator of seasonal sub-
aerial exposure in the lake margins (Carrión 2002),
although the paleocology of this organism remains
unclear (Scott 1992).
Figure 6. Pollen diagram with main taxa for core ZON-01-1A. (hatched pattern 5· exaggeration). Sedimentary facies associations are also shown. Dots indicate ‘presence’ with
percentages minor than 2%.
453
454
Only five samples were analyzed in core ZON-
01-1B to characterize the pollen assemblages of the
main units defined (Figure 4). The location of the
Olive rise (between 75 and 100 cm) provides a time
horizon to correlate both cores.
Pollen assemblages from aquatic and littoral
plants serve as paleohydrological indicators. Pollen
samples from Zone III contain the highest per-
centages of Myriophyllum spicatum and relatively
high of Potamogeton and Ruppia. Although Ruppia
is an indicator of moderate saline waters, the pollen
association suggest lake waters were fresher than
today, likely due to a higher fluvial input. In zone
II, taxonomic diversity is the highest, and pollen
percentages of plants from vegetated lake margins
(Tamarix, Sparganium) increase. Freshwater
aquatic plants progressively decrease and hygro-
phytes plants typical of littoral vegetation reach the
highest percentages at the top of this zone. At this
time, littoral vegetated areas surrounding the lake
would have reached the largest size and lake waters
would be chemically more concentrated. In zone I,
the decrease in hygrophyte pollen content and the
increase in Potamogeton and Myriophyllum is
interpreted as a reflection of the decreasing surface
area occupied by the vegetated littoral zone when
lake levels were relatively higher.
Diatoms
Diatom assemblages in core ZON-01-1A are
dominated by the tychoplanktonic Fragilaria
brevistriata, and the periphytic Cocconeis neot-
humensis throughout most of the core (Figure 7).
Some levels have also large numbers of the
planktonic Cyclotella meneghiniana and the peri-
phytic Amphora pediculus among others. Re-
worked marine diatom taxa include Chaetoceros
resting spores, Actinoptychus senarius, Thalassio-
nema nitzschioides, Asteromphalus sp. and Neo-
denticula sp. Their presence indicates periods of
increased sediment input to the lake from the
Miocene marine calcarenites, the dominant
lithology in the watershed. Five diatom assem-
blage zones (DAZs) have been defined (Figure 7).
DAZ ZON-V (161–141 cm)
This zone is dominated by the salinity – indifferent
(Tapia et al. 2003) euplanktonic Cyclotella me-
neghiniana and, towards the top of the zone, by the
freshwater tychoplanktonic Fragilaria brevistriata.
These taxa suggest an environment characterized
by open water conditions and moderate depth at
the coring site.
DAZ ZON-IV (141–111 cm)
Fragilaria brevistriata and the freshwater benthic
Amphora pediculus are the main taxa in this zone.
The co-dominance of the latter in the assemblage
points to a lowering of the lake level compared to
the previous DAZ.
DAZ ZON-III (111–71 cm)
Fragilaria brevistriata is still a co-dominating taxa
in this zone, but the oligosaline benthic Cocconeis
neothumensis becomes the predominant species,
which suggests, as in the previous zone, lower lake
levels, but of more saline conditions.
DAZ ZON-II (71–51 cm)
Both Fragilaria brevistriata and Cocconeis neot-
humensis show a decrease in this zone which is
dominated by the planktonic Cyclotella menegh-
iniana, suggesting a new increase in water levels.
DAZ ZON-I (51–30 cm)
This zone represents a substantial change in dia-
tom assemblages, which are now dominated by
benthic forms of both freshwater (mainly Cym-
bella microcephala and Fragilaria capucina var.
gracilis) and saline (mainly Cymbella affinis,
Cocconeis placentula and Nitzschia elegantula)
preferences. This zone also shows an important
increase in the allochthonous marine taxa such as
Chaetoceros spp. and Thalassionema nitzschioides.
The dominance of benthic taxa suggest lower lake
levels, but the mixture of freshwater and saline
forms also points to rapid changes in the contri-
bution of freshwaters. Flooding episodes could
explain not only those changes, but the increase
in the percentages of marine reworked taxa as
well.
The upper 30 cm of the core is considered a non-
countable interval. Some levels show no diatom
preservation, and the levels where diatoms were
preserved are almost 100% composed of non-
countable Thalassionema spp. small fragments.
This zone denotes an increase in the erosion
activity of the basin responsible for the input of
allochthonous diatoms to the lake from the Mio-
cene marine rocks.
Figure 7. Diatom percent abundance diagram of ZON-01-1A for taxa ‡2% (nci = non countable interval). Sedimentary facies associations are also shown.
455
456
Discussion
Results from the two studied cores provide a re-
cord of past climate, environmental and human-
induced changes in the Laguna Zoñar watershed
since early Medieval times, and also an opportu-
nity to test the impact of the restoration mea-
surements applied by the regional government
during the last 20 years. In Figure 8, we summa-
rize the proxy-records and compare them with the
main historical events and the available climate
reconstructions. We distinguish six main periods
of human and climatic interactions in the
watershed:
Period 6: prior to mid-13th century
This period is represented by sedimentary Unit 4,
composed of facies arranged in cm-thick sequences
that suggests relatively rapid changes in deposi-
tional environments in a lake highly influenced by
fluvial input, and with prevalent oxic conditions at
the bottom of the lake. The boundaries between
the three subunits could correspond to deposi-
tional hiati, however the only sedimentological
evidence for an unconformity occurs at the top of
the unit, where an erosive surface marks the base
of the next sequence. Chronological control for
this unit is provided by the basal AMS date
(165 cm, 1771 ± 3814C yr BP, ca. AD 300) and the
estimate age of the erosive surface at the top
(around 1250 AD). Most likely, the unit represents
deposition during the post-Roman and early Vis-
igoth Age; the sediments corresponding to the
Muslim Period have been eroded.
A strong fluvial influence in this sediment
interval is marked by coarser, more clastic facies
arranged in fining upward sequence, and the
presence of Potamocypris sp., a nektic ostracode
typical of littoral environments. The dominance of
planktonic diatoms and aquatic plants such as
Myriophyllum points to relatively higher lake levels
than today and probably fresh waters with low
chemical concentrations. The Arroyo de las Sali-
nas outlet was likely functional and lake level was
at the highest. The large flooding episode at the
base of Unit 3 (140 cm, facies 4, magnetic sus-
ceptibility peak) was more related to climatic fac-
tors (increase in precipitation) than to increased
disturbance of the watershed. Enhanced fluvial
activity has been documented in several Iberian
river basins during the cooler Holocene climate
events such as the Early Medieval Ice Advance (6–
10th century) and the Little Ice Age (15–19th
century), and less significant fluvial dynamics oc-
curred during the Medieval Climate Optimum
(Peña-Monné et al. 1998; Benito et al. 2003). The
10–11th century is generally characterized in Eur-
ope as one of the warmest historic periods (the so-
called Medieval Warm Period). In northern and
central Europe severe winters were somewhat less
frequent and less extreme during the MWP, AD
900–1300, than in the ninth century and from 1300
to 1900 (Pfister et al. 1998). In Spain, high flood
frequencies were registered in most of the Atlantic
watersheds during the late MWP (AD 1160–1210)
(Benito et al. 2003).
Probably, the regional vegetation during this
period comprised sclerophyllous formations of
evergreen Quercus, Olea europaea sylvestris,
Ceratonia siliqua, Pinus, Juniperus and Mediter-
ranean shrubs; mesohygrophytic vegetation with
deciduous trees in protected areas as humid gor-
ges, rivers (Alnus, Fraxinus, Populus, Ulmus) and
some herbaceous extensions. The high percentages
of Olea in the Zoñar record since the base indicate
that olive trees were a significant element of the
Mediterranean landscape. In diverse paleopaly-
nological studies from the Mediterranean area of
Iberia, a notable presence of Olea has been de-
tected since the last glaciation (Carrión 1992;
Burjachs and Julià 1994; Pérez-Obiol and Julià
1994; Carrión and van Geel 1999; Carrión 2002;
Pantaleón-Cano et al. 2003). As far as more recent
times are concerned, the presence of Olea is also
detected in northern regions of Iberia from the
Early Holocene (Davis 1994; Riera and Esteban
1994), and the mid-Holocene in the Balearic
islands (Yll et al. 1996, 1997).
Nevertheless, the expansion of Olea in historical
times is due to human intervention. The cultivated
olive was introduced in Spain by the Phoenicians
and Greeks, although wild olive was probably
used by the indigenous people. An increase in
pollen percentages around 5–10% appears in sev-
eral sites in Portugal (Van den Brink and Janssen
1985) and northeastern Spain (Saladas de Alcañiz:
Davis 1994; Estaña lake: Riera et al. 2004) coin-
ciding with the Roman Period. Various tree and
shrub taxa (Olea, Castanea, Juglans, Vitis, Fraxi-
nus, Platanus, etc.) are clearly cultivated from
Figure 8. Summary of the environmental and climate changes reconstructed from several proxy records in the Zoñar cores and correlation with the main historic events in the
watershed. Sedimentary facies and Units as in Figure 4. The rainfall index (Rodrigo et al. 1999, see text) is also indicated.
457
458
3000 yr BP onwards in the western Mediterranean
(Davis 1994; Sadori and Narcisi 2001; Pantaleón-
Cano et al. 2003). In Central Italy, cultivation of
Castanea, Juglans and Olea started at 3000–
2900 cal yr BP (Mercuri et al. 2002). The first
significant spread of Olea in Lago di Pergusa
(Sicily) occurs between 3300 and 3000 years BP
and it could be interpreted either as the beginning
of extensive cultivation or increasing arid condi-
tions; however, the maximum expansion of Olea
clearly related to human activities occurs between
2600 and 2400 years BP (Sadori and Narcisi 2001).
The Visigoth period (415–711 yr AD) is gener-
ally characterized by a progressive depopulation of
urban areas and lower agricultural use of the land.
However, some pollen records do not show chan-
ges in olive percentages, suggesting that areas al-
ready farmed during Roman times continued in
production (Van den Brink and Janssen 1985;
Riera et al. 2004). Changes in agricultural use of
the land were more significant during the Arab
period in southern Spain and particularly the
Guadalquivir River Valley. Lagoa Comprida in
Portugal (Van den Brink and Janssen 1985) shows
a sharp rise in olive percentages up to 20% during
this period. The city of Cordoba attained a pop-
ulation over 500,000, the largest and most pros-
perous city in Europe and the Mediterranean
Basin. Although there is no sediment record from
this period in the analyzed Zoñar cores, historical
documents show that the agricultural landscape of
the Campiña Cordobesa was dominated by olive
tree orchards since the Roman times.
Period 5: From the Christian conquest till the end
of the ‘golden age’ (13–18th centuries)
This period corresponds to sedimentary Unit 3
and it is characterized by two sequences (subunit
3B and 3A), both showing an upward trend to-
wards lighter colors, and increasing lacustrine
(authighenic and biogenic) component in the sed-
iments: gray, massive carbonate mud (F 3), mas-
sive brown (F 2), and finely laminated brown
sediments (F 1). These sedimentological features
point to reducing fluvial input into the lake and the
dominance of ‘lacustrine’ processes in the deepest
area were the core was retrieved. The decrease in
Myriophyllum spicatum and other aquatic plants
and the generally low values during this unit sug-
gest lower lake levels and more concentrated wa-
ters. A general trend to increased percentages of
benthic diatoms of saline condition supports this
interpretation.
A relatively small decrease in olive pollen occurs
at the base of this unit that could be related to
some of the events in the Guadalquivir Valley
during the 13–14th centuries, although the inter-
pretations must be cautious because the coarse
pollen and chronological resolution. The city of
Cordoba was conquered by the Christians in AD
1236, and the Guadalquivir River Valley was re-
populated during the 13th and 14th centuries.
However, till the conquest of the Kingdom of
Granada (1492), the region was the border be-
tween Christian and Muslim Kingdoms and mili-
tary incursions were frequent. The forest and fields
were usually burnt during the military fights, and
consequently, some farmland was abandoned and
farming substituted by sheep and goat husbandry.
The small decrease in pollen percentages at the
base of pollen zone II could be a reflection of the
agricultural regression during the 13th and 14th
centuries in the Guadalquivir Valley.
Historical data show three periods of acceler-
ated human impact on the landscape of the Cor-
doba Province characterized by increasing
cultivated land (Ortega Alba 1975): (i) between the
Christian conquest of the Guadalquivir River
valley and the 18th century, (ii) the 19th century
after Church-owned land was expropriated and
more intensively cultivated, and (iii) the mid-20th
century. The progressive increase in olive pollen in
Unit 3 is a reflection of the intensification of
agriculture in the Campiña Cordobesa since the
15th century. Once the border between Christian
and Muslim kingdoms was more secure and par-
ticularly after the fall of Granada, the Campiña
experienced a period of rapid transformation of
the landscape due to the new structure of the land,
the increase in agricultural exploitations, and a
higher olive production. This would be the ‘first’
olive rise that occurs in the record at about 125 cm
depth. Olive pollen doubled and maintained a
slightly increasing trend throughout this period.
Chemical profiles do not show an increase in soil
erosion; aluminum concentrations slightly decline
upcore as an evidence of smaller alluvial contri-
bution to the lake sediments.
The rise in olive cultivation started in the 14th
and 15th centuries throughout Spain, and contin-

ued during the 16–17th centuries characterized by
an increased agricultural intensification. In north-
eastern Spain, when the Christian conquest was
completed there are not references to olive groves,
but later, in 1300 AD, vineyards were replaced by
olive cultivations and the production of oil
reached a peak in the 14th century (Riera et al.
2004). Many olive curves from Spanish pollen sites
show the largest increase during the Holocene at
about the 16th century, particularly in central and
northeastern sites (Davis 1994; Riera et al. 2004).
In spite of some unfavorable climatic conditions
during the first part of the LIA, as frequent severe
frost during the winter in Córdoba and periods
when the Guadalquivir River froze over at the end
of the 16th century (Font 1988), pollen record in
Zoñar shows high values and a slight increase. The
significant decline observed in Lagoa Comprida at
about 500 yrs BP, interpreted as a decline in
agricultural productivity (Van den Brink and
Janssen 1985) and other periods of known social
unrest and population crises as the 15th century do
not show in the Zoñar record. However, the rela-
tively low resolution of our sampling precludes
any conclusion on the impact of such events on
olive production in the Campiña Cordobesa.
The Modern Age Period (17–18th centuries)
would correspond to subunit 3A (AD 1660–1760)
composed of another fining upward sequence with
gray, more clastic facies 4 at the base topped by
brown facies 1 and finely laminated facies. A
general decrease in fine particles (<2 microns)
indicates a reduction in clastic input to the lake
towards the top of this interval. Microscopic
observations indicate that the higher percentages
of large particles reflect the increase in diatom
content and not of coarse clastic particles. This
sequence represents another gradual transition
from a clastic-dominated lake subenvironment
with significant fluvial input, to a mixed clastic-
authigenic subenvironment. Aluminum content, a
chemical indicator of watershed erosion, decrease
during this interval. Iron content also decreases
although because of the higher chemical mobility
of this element it is not a straightforward indicator
of catchment erosion. Higher values of Cyclotella
meneghiniana (DAZ ZON-IV) suggest another
episode of relatively higher lake levels at the top of
Unit 3.
Climate reconstructions in Andalucı́a based on
documentary data and instrumental measurements
459
show that changes in the rainfall regime have been
more important than those in the temperature
during the last centuries (Rodrigo et al. 1999). A
rainfall index for the last 500 years has been cal-
culated by Rodrigo et al. (1999, 2000) based on
documentary and instrumental sources for the
period 1791–1997 (Figure 8). Seasonal indexes
derived from documentary sources were assigned
as follows: 2 (Very dry, absence of rain together
with information on drops in river and water
spring levels), 1 (Dry, absence of rain without
additional information), 0 (Normal, no news, no
comments on impacts), +1 (Wet, hard and/or
constant rainfall), +2 (Very wet, rain-induced
flood). The index values derived from instrumental
sources for the period 1791–1997 are expressed as
follows (R = rainfall; Ri: ith-percentile of the
instrumental series): Very dry: R < R10; Dry:
R10 < R < R25; Normal: R25 < R < R75; Wet:
R75 < R < R90; Very wet: R > R90. The Annual
index is the sum of the four seasonal indices and its
value ranges from 8 to +8, although lack of data
relating to the summer prevents such extreme
values to be reached. The results suggest a fluctu-
ating evolution since the end of the Medieval
Warm Period, with the wettest intervals during the
late 16–mid-17th centuries (AD 1590–1650) and at
the end of the 19th century. The driest periods
occurred during the first half of the 16th century
(AD 1501–1589) and during the AD 1650–1750
period, and a general trend to more arid condi-
tions started in the early 20th century (Rodrigo
et al. 1999, 2000).
The rainfall index must be interpreted as a
measure of the behavior of weather anomalies or
extreme phenomena, rather than average values,
because it is based on historical documents
reporting only exceptional socio-economical im-
pacts. Although our chronology is not accurate
enough to compare with these climatic fluctua-
tions, some tentative correlations can be estab-
lished. The wettest interval (late 16–mid-17th
centuries; AD 1590–1650) could correspond to the
top of Unit 3B and the transition to 3A. In subunit
3A, the presence of more clastic facies 4 at the base
of a fining-upward sequence suggest increased
fluvial activity in the creeks, although aquatic
pollen and diatoms do not indicate a particularly
high lake level phase. Sedimentological features,
aquatic pollen and diatom assemblages mark the
top of Unit 3 (subunit 3A) as a period of higher
460
lake levels that could correspond to a secondary
wet period between the late 17th and the early 18th
century. In another Andalucian lake, the Laguna
de Archidona (Málaga Province), a dry period
during the mid-17th century dried out the lake and
some weak soil formation processes took place
(Luque et al. 2004); during the following period
between AD 1650–1850 lake level rose, forest
recovered, and agricultural practices were favored
instead of grazing. Documentary sources also
indicate the presence of larger shallow lakes in the
Doñana National Park prior to the 19th century
(Sousa and Garcı́a-Murillo 2003).
Period 4: the late 18th–early 19th century period
A large limnological change occurred in the lake
during the late 18th–early 19th century prior to the
sharp increase in olive pollen. Better-defined lami-
nation and progressive dominance of brown lami-
nae in subunits 2G and 2F indicate that suitable
conditions for the establishment of benthic bacte-
rial and algal communities at the bottom of the
lake were reached. These communities were dis-
turbed by deposition of two cm-thick layers of fine
sediments (Facies 5) that reflect two large flooding
episodes in the lake (Subunit E). Thin (about 1 mm
thick) clastic, gray laminae also occur at the base of
subunit D, and consequently, fine grain percent-
ages remain high. The mixture of diatoms of both
freshwater and saline character in DAZ ZON-I
could be an indication of the short-term changes in
salinity as an effect of the freshwater flooding epi-
sodes. A brackish to saline lake, with anoxic bot-
tom waters that prevented bioturbation and
facilitated the development of bacterial and algal
benthic communities was established during this
time. The low clastic input and the higher salinity
suggest generally lower lake levels than during
previous units. Both, the decrease in Myriophyllum
values and increase in Sparganium in Unit 2 point
to a period of more concentrated waters and rela-
tively lower lake levels with increasing littoral
vegetation. This enhanced macrophytic develop-
ment is also corroborated by the increase in benthic
diatoms of an epiphytic condition. The lake prob-
ably experienced a significant siltation as indicated
not only by the dominance of benthic diatoms but
also by the increase in tube-dwelling diatom forms
such as Cymbella (Figure 7).
Wet anomalies also dominated between AD
1750 and 1850 (Rodrigo et al. 1999, 2000) a period
that, according to our chronological model would
correspond to the deposition of the lower part of
Unit 2 (G, F, E and D subunits) characterized by
laminated sedimentary facies, biological evidence
for decreasing lake levels, and the presence of
clastic facies indicative of floods in the lake wa-
tershed (particularly in subunit E).
Period 3: 19th century (olive rise) till mid-20th
century (introduction of modern farm machinery)
Subunit C contains the best-laminated, organic-
rich facies of the core (Facies 7). During this per-
iod, lake level remained lower than present, fluvial
input was low, and chemical concentration in-
creased. Benthic diatoms dominate the assem-
blages (DAZ ZON-I), aquatic plants abundances
decrease and the littoral vegetated area with
Sparganium increases. Rainfall reconstructions
during the 19th century show wet anomalies dur-
ing the early and mid-decades, dry anomalies
during the 1860–1880s, and the onset of a dry
trend from the late 19th century (Figure 8). This
period of more arid climatic conditions during the
late 19th and the early 20th century corresponds to
the deposition of the organic, finely laminated fa-
cies of subunit 2C to 2A indicative of the lowest
lake level in the cores. Other wetlands and lakes in
Andalucı́a show evidence of a more negative
hydrological balance after the mid 19th century.
Reduction of surface area occupied by shallow
lakes in the Doñana National Park at the end of
the LIA indicate an increase in aridity (Sousa and
Garcı́a-Murillo 2003) that could correlate with the
onset of unit 2 in the Zoñar records. In the Laguna
Archidona, located in the Málaga Province, a
transition from laminated sediment to increasingly
gypsum-rich sediments occurs after AD 1850 and
it is interpreted as increasing aridity after the end
of the LIA (Luque et al. 2004).
Since the end of the 19th century, a progressive
decrease in rainfall took place, only interrupted by
a relatively wet period during the 1960s (Figure 4).
Increased human use of the water for irrigation
could have also helped to lowering lake level
during this period. Several historical documents
provide some information on past lake level in
Laguna Zoñar. In the mid-19th century Madoz

(1850) described Laguna Zoñar and reported two
springs and one creek feeding the lake. He noted
the depth of Laguna Zoñar as 34 ‘varas’ (1 ‘vara’
equals about 84 cm, so the depth was about
28.5 m). Although it seems unlikely that the lake
was that deep, Madoz’s description indicates that
the lake level was high during the mid-19th cen-
tury. Reconstructed rainfall for the late 19th cen-
tury also shows some positive anomalies (Rodrigo
et al. 1999). Dantin (1940) estimated the size as
2000 · 250 m and recorded that the surface outlet
was functional and that Laguna Chica existed as a
different lake.
The onset of an increasing trend in Fe and Al
concentrations in the sediments marks the begin-
ning of a period of significant soil erosion in the
basin. Olive pollen sharply increases. All these
indicators point to large human disturbance in the
Zoñar watershed during the mid and late 19th
century. The agrarian crisis of the 18th century did
not have an impact on the continued expansion of
olive in Spain. On the contrary, many records
show that olive production in Spain peaked in the
18th and 19th centuries. Some regions were dedi-
cated to specific crops required by industry, such
as hemp, though there were large areas of olives
and cereals and some regions expanded its olives
groves and doubled their oil production (Davis
1994; Riera et al. 2004). Several factors may have
helped: expropriation of church in 1837 brought
more land for intensive cultivation, changes in the
pattern of land ownership, and reduced frequency
of winter frost at the end of the LIA. The sharp
increase in olive pollen in Zoñar likely occurred
during the late 19th century. Several laws signed
by the Spanish Government during the late 19th
and early 20th century favored the drainage of
wetlands and most likely, the deepening and
drainage of the Arroyo de las Salinas occurred at
that time.
Subunit 2A is composed of variegated, lami-
nated Facies 7 that become more irregular and
dominated by gray laminae towards the top. It is a
transitional interval between the laminated facies
of Unit 2 and the massive facies of Unit 1. Organic
matter values decrease towards the top, till values
below 10%. Although the upper part of this sub-
unit is still laminated, a clear change occurs in the
composition of the sediments: higher clastic input
is marked by the increase in finer particles and in
magnetic susceptibility values. A decrease in large
461
particles (bioclasts and diatoms) and the disap-
pearance of green and brown laminae stresses
another abrupt limnological change that antici-
pated the onset of Unit 1. Although lake levels
remain low, conditions are not longer conducive to
the development of bacterial mats.
Period 2: mid-20th century to declaration
of the lake as protected area (1982)
The base of Unit 1 (Subunit 1C) still shows some
faint lamination (Facies 2). Dark gray, massive
sediments (Facies 4) with high magnetic suscepti-
bility values and low organic content constitute
subunit B. This period correlated with the intro-
duction of machinery around the mid-20th century
that provoked a rapid increase in soil erosion in
the watershed. Aluminum and iron concentration
sharply increase. Reworked marine diatoms dom-
inate as a consequence of intense erosion of the
Miocene marine bedrock in the watershed. Olive
pollen maintains high values, and the decrease in
Olea at the base of Unit 1 may be a reflection of
increased littoral vegetation around the lake more
than to changes in agricultural practices. Lake
levels are still low as indicated by the extension of
hygrophytes (high Cyperaceae and Poaceae val-
ues). Geochemical indicators show that this is the
period of most intense human disturbance of the
catchment and the lake hydrology. In the 1960s,
the waters from the Zoñar and Escobar springs
were diverted for human use to the nearby Aguilar
de la Frontera town, lake level lowered, and the
outlet creek Arroyo de las Salinas became non-
functional. The width of the littoral vegetation
zone surrounding the lake, increased. In the late
1970s water diversion for human consumption
progressively stopped and lake levels begin to
recover. Farming activities in the watershed
remained intense, and consequently erosion as
detected by Fe and Al profiles increased. The in-
crease in copper reflects the increasing use of fer-
tilizers since the 1960s. A general decreasing trend
in precipitation is observed after 1960 (Rodrigo
et al. 1999). A dry period during the early 1970s
could correlate with the deposition of faintly
laminated, brown sediments with higher organic
matter content in subunit 1A. Increasing grain size
at the top of the core is related to the abundance of
large oxidized organic matter remains, soil
462
particles and littoral plant remains, and it parallels
a large increase in magnetic susceptibility.
Period 1: restoration (since 1982)
After the lake was declared a protected area in
1982, the spring waters were not diverted for
human consumption and farming stopped in some
fields bought by the regional government. Average
lake level recovered and some littoral areas were
submerged. The Laguna Chica was flooded again,
and re-connected with the Laguna Zoñar, but the
outlet creek did not reopen; currently, Laguna
Zoñar does not have a surface outlet.
The top sample analyzed in the core is repre-
sentative of the modern status of the lake system.
The Al content is smaller compared to previous
values and magnetic susceptibility also decreases,
which suggest that reducing the surface of farming
has helped a little to alleviate the erosion rate in
the basin. Iron and phosphorous contents have
also declined. Olive pollen maintains the same
values. Copper remains high, which suggest that
the total amount of fertilizers reaching the lake
have not substantially changed. Myriophyllum
spicatum appears again as an indicator or rela-
tively higher lake levels and fresher waters. The
decrease in Poaceae pollen seems to reflect the
decrease of the littoral vegetation, now partially
submerged.
Since 1989 the Andalucian Environmental
Agency has started a restoration program with
autochthonous species as Quercus ilex, Quercus
coccifera, Olea europaea var. sylvestris, Ceratonia
siliqua, Populus alba, Tamarix gallica, Ficus carica,
Celtis australis, Crataegus monogyna, Pistacia
lentiscus, Arbutus unedo, Viburnum tinus, Retama
sphaerocarpa, Myrtus communis, etc. The shoreline
of the Laguna Zoñar is vegetated with Phragmites
australis and Typha dominguensis while Juncus
maritimus and Tamarix canariensis develop in the
waterlogged littoral areas. Dry shorelines are
covered with Polypogon maritimus and Plumba-
ginaceae as Limonium echioides. Zannichellia pa-
lustris is the main submerged plant. All these
species appear in the modern pollen rain samples
and also at the top samples of the cores. The
natural vegetation would be characterized by
sclerophyllous formations with some patches of
mesohygrophytic vegetation very restricted spa-
tially and altered in their composition, although
some taxa are present in the Zoñar area.
Conclusions
The study of two sediment cores from Laguna
Zoñar (Andalucı́a, Spain) provides a detailed
record of environmental, climatic and anthropo-
genic changes in a Mediterranean watershed since
Medieval time. The direct relationship between
rainfall and lake level observed during the last
decades suggests that climate variability is a main
controller of lake level in the past. Sedimentolog-
ical and biological proxies indicate that higher lake
levels dominated prior to the 13th century. En-
hanced fluvial influence at the end of the Medieval
Warm Period (ca. AD 1300) could be responsible
for some erosion at the coring site and the gener-
ation of an erosive hiatus. There is not a direct
correlation between rainfall anomalies recon-
structed for the last 500 years from documentary
records and the inferred lake level changes in Zo-
ñar (Figure 8). This may be due to the fact that the
rainfall index only reflects extreme events and the
lake filters and smooth the climatic signal and also
to the uncertainty of our chronological model. The
most significant limnological change started in the
late 18th century where more finely laminated fa-
cies deposited and it corresponds to a period of
dominant wet anomalies. A dry period at the end
of the 19th century corresponds to the onset of
deposition of finely laminated, organic – rich facies
during a low lake level stage. This dry phase at the
end of the LIA has been identified in historical
documents and also in other lake records in the
region. The Zoñar record shows fluctuating lake
levels since the end of the Medieval Warm Period
till the late 19th century and a more acute dry
period during the late 19th century – early 20th
century, after the end of the Little Ice Age. This is
in agreement with historical records document
high climate variability during the 14–19th century
in the Iberian Peninsula, with periods of intense
rainfall and droughts (Rodrigo et al. 1999, 2000)
and with dendroclimatic reconstructions that show
outstanding oscillations during the LIA (Manri-
que and Fernández-Cancio 2000). Although the
onset to lower lake levels characteristic of Unit 2
does not correlate with a dry rainfall anomaly, the
deposition of finely laminated, organic-rich facies

correlate with the arid period during the late 19th
century (end of the LIA) identified in documentary
records and also in several other sites in the region.
Water consumption for human use and farming
during the late 19th and early-mid-20th century
likely intensified this trend.
Two main periods of increased human activities
in the watershed are recorded in the sediments.
The first started with the Christian conquest and
colonization of the Guadalquivir River Valley
(13th century) particularly after the fall of the
Granada Kingdom (15th century). The second one
corresponds to the late 19th century when more
land was cultivated after expropriation of the
Church. Intensification of soil erosion occurred in
the mid-20th century, after farm machinery was
introduced. The 137Cs chronology indicates a very
large increase in sedimentation rate during the last
decades, when massive calcite mud deposited (Unit
1) on top of the variegated, laminated mud of Unit
2. Human activities may have played a role in
Zoñar hydrology changes since the late 19th cen-
tury, when the outlet was drained, and particularly
in the mid-20th century (till 1982) when the waters
were diverted for human use.
The end of deposition of laminated facies star-
ted at about 1960 and does not correlate with a
significant change in rainfall. Increased farming
activity may have played a major role in this lim-
nological change in the lake. Once the lake was
declared a protected area in the early 1980s, the
average lake levels increased. Pollen indicators
reflect this limnological change during the last few
decades. Geochemical indicators show a relative
decrease in soil erosion during the last decades, but
no change in the amount of fertilizers reaching the
lake. Our study also provides an opportunity to
evaluate the relative significance of human versus
climatic factors in lake hydrology and watershed
changes during historical times. These paleolim-
nological reconstructions can be used by natural
resources agencies to better define the lake man-
agement policies and to assess the results of the
restoration efforts started two decades ago.
Acknowledgements
Financial support for research at Laguna Zoñar
was provided by the Spanish Inter-Ministry
Commission of Science and Technology (CICYT),
463
REN 2000–1136 Project ‘‘Arid periods in the
Mediterranean areas of the Iberian Peninsula since
the last glacial maximum: chronology, character-
ization and paleoclimate implications’’. Funding
by projects REN 2003–02499GLO and PI 17/739/
FS is also acknowledged. We thank the Director
and the staff of the ‘‘Southern Cordoba Lakes
Natural Reserve’’ for their help during fieldwork
and also for access to unpublished data. Dirk
Verschuren (Gent University, Belgium) helped to
retrieve and sample the cores. We are grateful to
Piero Guilizzoni and Bill Last for their comments
and criticisms that improved the manuscript.
References
Benito G., Dı́ez-Herrero A. and de Villalta M. 2003. Magnitude
and frequency of flooding in the Tagus Basin (Central Spain)
over the last millennium. Climate Change 58: 171–192.
Boyle J. 2001. Inorganic geochemical methods in palaeolim-
nology. In: Last W.M. and Smol J.P. (eds), Tracking Envi-
ronmental Change Using Lake Sediments, vol 2: Physical and
Geochemical Methods. Kluwer Academic Publishers,
Dordrecht, The Netherlands, pp. 83–141.
Boyle J. 2004. A comparison of two methods for estimating the
organic matter content of sediments. J. Paleolimnol. 31: 125–
127.
Bradbury J.P., Colman S. and Reynolds R.L. 2004. The history
of recent limnological changes and human impact on Upper
Klamath Lake, Oregon. J. Paleolimnol. 31: 151–165.
Burjachs F. and Julià R. 1994. Abrupt climatic changes during
the last glaciation based on pollen analyses of the Abric
Romanı́, Catalonia, Spain. Quaternary Res. 42: 308–315.
Buurman P., Pape Th. and Muggler C.C. 1997. Laser grain-size
determination in soil genetic studies: I. Practical problems.
Soil Sci. 162: 211–218.
Capel Molina J.J. 1981. Los climas de España. Oikos Tau
Ediciones, Barcelona, pp. 429
Carrión J. 1992. Late Quaternary pollen sequence from Cari-
huela cave, southeastern Spain. Rev. Palaeobot. Palynol. 71:
37–77.
Carrión J. 2002. Patterns and processes of Late Quaternary
environmental change in a montane region of southwestern
Europe. Quaternary Sci. Rev. 21: 2047–2066.
Carrión J.S. and Navarro C. 2002. Cryptogam spores and
other non-pollen microfossils as sources of palaeoecological
information: case-studies from Spain. Ann. Bot. Fenn. 39:
1–14.
Carrión J. and van Geel B. 1999. Fine-resolution Upper
Weichselian and Holocene palynological record fron Nava-
rrés (Valencia, Spain) and a discussion about factors of
Mediterranean forest sucession. Rev. Palaeobot. Palynol.
106: 209–236.
Carrión J., Andrade A., Bennet K., Navarro C. and Munuera
M. 2001. Crossing forest thresholds: inertia and collapse in a
Holocene sequence from south-central Spain. The Holocene
11: 635–653.
464
Carrión J., Yll E., Walker M., Legaz A., Chaı́ns C. and López
A. 2003. Glacial refugia of temperate, Mediterranean and
Ibero-North African flora in south-eastern Spain: new evi-
dence from cave pollen at two Neandertal man sites. Global
Ecol. Biogeogr. 12: 119–129.
Cohen A.S., Palacios-Fest M.R., Msaky E.S., Alin S.R.,
McKee B., O’Reilly C.O., Dettman D.L., Nkotagu H. and
Lezzar K.E. 2005. Paleolimnological investigactions of
anthropogenic environmental change in Lake Tanganyika:
IX. Summary of paleorecords of environmental change and
catchement deforestation at Lake Tanganyika and impacts
on the Lake Tanganyika ecosystem. J. Paleolimnol. 34: 125–
145.
Coshell L. and Rosen M. 1994. Stratigraphy and Holocene
history of Lake Hayward, Swan Coastal Plain Wetlands,
western Australia. In: Renaut R. and Last W. (eds), Sedi-
mentology and Geochemistry of Modern and Ancient Saline
Lakes. SEPM, Tulsa, Sp. Publ. 50, pp. 173–188.
CSIC 1976. Comisión de métodos analı́ticos. An. Edafol. Ag-
robiol. 35: 813–814.
Dantin J. 1940. La aridez y el endorreismo en España. El en-
dorreismo bético. Estudios geográficos 1: 75–117.
Davis B. 1994. Paleolimnology and Holocene environmental
change from endorheic lakes in the Ebro Basin, north-east
Spain. Ph. D. Thesis. University of Newcastle Upon Tyne,
pp. 317
Enadimsa 1989. Estudio hidrogeológico de la Laguna de Zoñar.
Junta de Andalucı́a, Agencia de Medio Ambiente, pp. 125
Font I. 1988. Historia del clima en España. Cambios climáticos
y sus causas. Instituto Nacional de Meteorologı́a, Madrid,
pp. 297
Forester R.M. 1988. Non marine calcareous microfossil sample
preparation and data acquisition procedures. United States
Geological Survey Technical Procedure HP-78 RI: 1–9.
Giralt S., Burjachs F., Roca J.R. and Julià R. 1999. Late
Glacial to Early Holocene environmental adjustment in the
Mediterranean semi-arid zone of the Salines playa-lake
(Alicante, Spain). J. Paleolimnol. 21: 449–460.
González-Sampériz P. 2004. Evolución paleoambiental del
sector central de la cuenca del Ebro durante el Pleistoceno
superior y Holoceno. Instituto Pirenaico de Ecologı́a-CSIC,
Zaragoza, pp. 210
Heiri O., Lotter A.F. and Lemcke G. 2001. Loss on ignition as
a method for estimating organic and carbonate content in
sediments: reproductibility and comparability of results. J.
Paleolimnol. 25: 101–110.
Hutchinson S.M. 2005. The recent sedimentation history of
Aqualate Mere (central England): assessing the potential for
lake restoration. J. Paleolimnol. 33: 205–228.
IGME 1988. Mapa Geológico de España, 1: 50000, (serie
MAGNA). Hoja 988. Puente Genil.Servicio de Publicaciones,
Ministerio de Industria y Energı́a, Madrid. 1 map, pp. 50
Kienel U., Schwalb M.J. and Schettler G. 2005. Distinguishing
climatic from direct anthropogenic influences during the last
400 years in varved sediments from Lake Holzmaar (Eifel,
Germany). J. Paleolimnol 33: 327–347.
Luque J.A., Julià R., Riera S., Marqués M.A., López-Sáez J.A.
and Mezquita F. 2004. Respuesta sedimentológica a los
cambios ambientales de épocas históricas en el sur de la
Penı́nsula Ibérica: La secuencia de la Laguna Grande de
Archidona (Málaga). Geotemas 6: 113–116.
Madoz P. 1850. Diccionario geográfico-estadı́stico-histórico de
España y sus posesiones de Ultramar. Madrid.
Manrique E. and Fernández-Cancio A. 2000. Extreme climatic
events in dendroclimatic reconstructions from Spain. Cli-
matic Change 44: 123–128.
Mercuri A.M., Accorsi C.A. and Bandini-Mazzanti M. 2002.
The long history of Cannabis and its cultivation by the Ro-
mans in central Italy shown by pollen records from Lago di
Albano and Lago di Nemi. Veg. Hist. Archaebot. 11: 263–
276.
Moore P., Webb J.A. and Collinson A. 1991. An Illustrated
Guide to Pollen Analysis. Hodder and Stroughton, London,
pp. 216
Moya J.L. 1984. Hidrogeologı́a de la Laguna de Zoñar. Oxyura
1: 21–41.
Moya J.L. 1986. Análisis del hidrograma del manantial de
Zoñar. Oxyura 3: 29–33.
Newton M. 1994. Holocene fluctuations of Mono Lake, Cali-
fornia: The sedimentary record. In: Renaut R. and Last W.
(eds), Sedimentology and Geochemistry of Modern and
Ancient Saline Lakes. SEPM, Tulsa, Sp. Publ. 50, pp. 143–
158.
Ortega Alba F. 1975. El Sur de Córdoba. Estudio de Geograf ı́a
Agraria. Tomo 2. Caja de Ahorros de Córdoba, Córdoba,
pp. 258
Pantaleón-Cano J., Yll E., Pérez-Obiol R. and Roure J.M.
2003. Palynological evidence for vegetational history in semi-
arid areas of the western Mediterranean (Almerı́a, Spain).
The Holocene 13: 109–119.
Peña-Monné J.L., Julián A., Chueca J. and Echeverrı́a M.T.
1998. Los estudios geoarqueológicos en la reconstrucción del
paisaje. Su aplicación en el valle bajo del rı́o Huerva (De-
presión del Ebro). Arqueologı́a Espacial. Arqueologı́a del
Paisaje 19–20. Teruel: 169–183.
Peñalba M.C., Arnold M., Guiot J., Duplessy J.C. and de
Beaulieu J.L. 1997. Termination of the Last Glaciation in the
Iberian Peninsula inferred from the Pollen sequence of
Quintanar de la Sierra. Quaternary Res. 48: 205–214.
Pérez-Obiol R. and Julià R. 1994. Climatic changes on the
Iberian Peninsula recorded in a 30,000-yr pollen record from
Lake Banyoles. Quaternary Res. 41: 91–98.
Pfister C., Schwarz-Zanetti G., Wegmann M. and Luterbacher
J. 1998. Winter air temperature variations in western Europe
during the Early and High Middle Ages (AD 750–1300). The
Holocene 8: 535–552.
Pons A. and Reille M. 1988. The Holocene and upper Pleisto-
cene pollen record from Padul (Granada, Spain): a new
study. Palaeogeogr. Palaeoclimatol. Palaeoecol. 66: 243–263.
Renaut R. and Tiercelin J.J. 1994. Lake Bogoria, Kenya Rift
Valley – a sedimentological overview. In: Renaut R. and Last
W. (eds), Sedimentology and Geochemistry of Modern and
Ancient Saline Lakes. SEPM, Tulsa, Sp. Publ. 50, pp. 101–123.
Renberg I. 1990. A procedure for preparing large sets of diatom
slides from sediment cores. J. Paleolimnol. 4: 87–90.
Riera S. and Esteban A. 1994. Vegetation history and human
activity during the last 6000 years on the central Catalan
coast (northeastern Iberian peninsula). Veg. Hist. Archaeo-
bot. 3: 7–23.
Riera S., Wansard G. and Julià R. 2004. 2000-year environ-
mental history of a karstic lake in the Mediterranean Pre-
Pyrenees: the Estanya lakes (Spain). Catena 55: 255–370.

Rodrigo F.S., Esteban-Parra M.J., Pozo Vázquez D. and
Castro-Dı́ez Y. 1999. A 500-year precipitation record in
southern Spain. Int. J. Climatol. 19: 1233–1253.
Rodrigo F.S., Esteban-Parra M.J., Pozo Vázquez D. and
Castro-Dı́ez Y. 2000. Rainfall variability in southern Spain
on decadal to centennial time scales. Int. J. Climatol. 20: 721–
732.
Sadori L. and Narcisi B. 2001. The Postglacial record of envi-
ronmental history from Lago di Pergusa, Sicily. The Holo-
cene 11: 655–670.
Sánchez M., Fernández-Delgado C. and Sánchez-Polaina F.J.
1992. Nuevos datos acerca de la morfometrı́a y batimetrı́a de
la Laguna de Zoñar (Aguilar de la Frontera, Córdoba).
Oxyura 6: 73–77.
Santisteban J.I., Mediavilla R., López-Pamo E., Dabrio C.J.,
Ruiz Zapata M.B., Gil Garcı́a M.J., Castaño S. and Martı́-
nez-Alfaro P.E. 2004. Loss on ignition: a qualitative or
quantitative method for organic matter and carbonate min-
eral content in sediments. J. Paleolimnol. 32: 287–299.
Schnurrenberger D., Russel J. and Kelts K. 2003. Classification
of lacustrine sediments based on sedimentary components. J.
Paleolimnol. 29: 141–154.
Scott L. 1992. Environmental implications and origin of
microscopic Pseudoschizaea Thiergart and Franz ex R. Po-
tonié emend in sediments. J. Biogeogr. 19: 349–354.
Sousa A. and Garcı́a-Murillo P. 2003. Changes in the westlands
of Andalusia (Doñana Natural Park, SW Spain) at the end of
the Little Ice Age. Climatic Change 58: 193–217.
Tapia P.M., Fritz S.C., Baker P., Seltzer G.O. and Dunbar R.
2003. A Late Quaternary diatom record of tropical climatic
history from Lake Titicaca (Peru and Bolivia). Palaeogeogr.
Palaeoclimatol. Palaeoecol. 194: 139–164.
Tylmann W. 2005. Lithological and geochemical record of
anthropogenic changes in recent sediments of a small lake
(Lake Pusty Staw, northern Poland). J. Paleolimnol. 33: 313–
325.
Van Den Brink L.M. and Janssen C.R. 1985. The effect of
human activities during cultural phases on the development
of montane vegetation in the Serra de Estrela, Portugal. Rev.
Palaeobot. Palynol. 44: 193–202.
465
Valero-Garcés B.L., Zeroual E. and Kelts K. 1998. Arid phases
in the western Mediterranean region during the Last Glacial
Cycle reconstructed from lacustrine records. In: Benito G.,
Baker V.R. and Gregory K.J. (eds), Paleohydrology and
Environmental Change. Wiley and Sons, London, pp. 67–80.
Valero-Garcés B.L., Navas A., Machin J., Stevenson T. and
Davis B. 2000a. Responses of a saline lake ecosystem in semi-
arid regions to irrigation and climate variability. The history
of Salada Chiprana, Central Ebro Basin, Spain. Ambio 26:
344–350.
Valero-Garcés B., González-Sampériz P., Delgado-Huertas A.,
Navas A., Machı́n J. and Kelts K. 2000b. Late Glacial and
Late Holocene environmental vegetational change in Salada
Mediana, central Ebro Basin, Spain. Quatern. Int. 73/ 74: 29–
46.
Valero-Garcés B., Navas A., Mata P., Delgado-Huertas A.,
Machı́n J., González-Sampériz P., Schwalb A., Ariztegui D.,
Schnellmann M., Bao B. and González-Barrios A. 2003.
Sedimentary facies analysis in lacustrine cores: from initial
core descriptions to detailed palaeoenvironmental recon-
struction. A case study from Zoñar Lake (Córdoba province,
Spain). In: Valero-Garcés B. (ed.), Limnogeology in Spain: A
Tribute to Kerry Kelts. CSIC, Madrid, pp. 385–414.
Valero-Garcés B., González-Sampériz P., Navas A., Machı́n
J., Delgado-Huertas A., Peña-Monné J.L., Sancho C.,
Stevenson A. and Davis B. 2004. Palaeohydrological
fluctuactions and steppe vegetation at the last Glacial
Maximum in the Central Ebro valley (NE Spain). Quatern.
Int. 122: 43–55.
Yll E., Pantaleón-Cano J., Pérez-Obiol R. and Roure J.M.
1996. Importancia de Olea en el paisaje vegetal del litoral
mediterráneo durante el Holoceno. In: Ramil Rego P.,
Fernández-Rodrı́guez C. and Rodrı́guez Guitián M. (eds),
Biogeografı́a Pleistocena-Holocena de la Penı́nsula Ibérica.
Consellerı́a de Cultura, Xunta de Galicia, Santiago de
Compostela, pp. 116–134.
Yll E., Pérez-Obiol R., Pantaleón-Cano J. and Roure J.M.
1997. Palynological evidence for climatic change and human
activity during the Holocene in Minorca (Balearic Islands).
Quatern. Res. 48: 339–347.